A model of neuromuscular synaptogenesis predicts that factors released by the nerve, such as agrin, determine the time and site of AChR clustering. These factors are thought to bind to specific receptors on the muscle fiber and activate intracellular signaling pathways, which act ultimately on the AChR, on the postsynaptic cytoskeleton, or on proteins, such as the 43K protein, that link AChR to the cytoskeleton. This model is not yet entirely proven, but provides a clear framework for further studies. Clustering of sodium channels at the neuromuscular postsynaptic membrane may also involve interaction with the cytoskeleton, but the signaling molecules are not yet known. AChR clustering and sodium channel clustering are separable events, both in the molecules involved and in developmental timing. Ion channel clustering in the CNS is less well understood, but may have features in common with the neuromuscular model. For example, glycine receptors are associated with gephyrin, a protein that mediates linkage to synaptic microtubules, thus raising obvious analogies to muscle AChR and the 43K protein. It is unknown, however, if AChR clustering in neurons is regulated by mechanisms in common with muscle. Coclustering of ion channels at synapses, both across the synaptic gap and within the same postsynaptic site, is likely to be of major importance to synaptic modulation and integration. Alignment of postsynaptic receptors with presynaptic calcium channels is critical for rapid transmission, but is achieved by mechanisms not yet understood. According to current models, the codistribution of two types of glutamate receptors in the postsynaptic membrane on dendritic spines is important for initiation of LTP. It is reasonable to expect that the basic mechanisms currently being elucidated at the neuromuscular junction will lead to greater understanding of how the CNS uses ion channel distribution to modulate synaptic activity.